Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “OpenCL”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 records

Evaluation of the OpenCL AES Kernel using the Intel FPGA SDK for OpenCL

The OpenCL standard is an open programming model for accelerating algorithms on heterogeneous computing system. OpenCL extends the C-based programming language for developing portable codes on different platforms such as CPU, Graphics processing units (GPUs), Digital Signal Processors (DSPs) and Field Programmable Gate Arrays (FPGAs). The Intel FPGA SDK for OpenCL is a suite of tools that allows developers to abstract away the complex FPGA-based development flow for a high-level software development flow. Users can focus on the design of hardware-accelerated kernel functions in OpenCL and then direct the tools to generate the low-level FPGA implementations. The approach makes the FPGA-based development more accessible to software users as the needs for hybrid computing using CPUs and FPGAs are increasing. It can also significantly reduce the hardware development time as users can evaluate different ideas with high-level language without deep FPGA domain knowledge. In this report, we evaluate the performance of the kernel using the Intel FPGA SDK for OpenCL and Nallatech 385A FPGA board. Compared to the M506 module, the board provides more hardware resources for a larger design exploration space. The kernel performance is measured with the compute kernel throughput, an upper bound to the FPGA throughput. The report presents the experimental results in details. The Appendix lists the kernel source code.

97 MATHEMATICS AND COMPUTING↗

Evaluation of CHO Benchmarks on the Arria 10 FPGA using Intel FPGA SDK for OpenCL

The OpenCL standard is an open programming model for accelerating algorithms on heterogeneous computing system. OpenCL extends the C-based programming language for developing portable codes on different platforms such as CPU, Graphics processing units (GPUs), Digital Signal Processors (DSPs) and Field Programmable Gate Arrays (FPGAs). The Intel FPGA SDK for OpenCL is a suite of tools that allows developers to abstract away the complex FPGA-based development flow for a high-level software development flow. Users can focus on the design of hardware-accelerated kernel functions in OpenCL and then direct the tools to generate the low-level FPGA implementations. The approach makes the FPGA-based development more accessible to software users as the needs for hybrid computing using CPUs and FPGAs are increasing. It can also significantly reduce the hardware development time as users can evaluate different ideas with high-level language without deep FPGA domain knowledge. Benchmarking of OpenCL-based framework is an effective way for analyzing the performance of system by studying the execution of the benchmark applications. CHO is a suite of benchmark applications that provides support for OpenCL [1]. The authors presented CHO as an OpenCL port of the CHStone benchmark. Using Altera OpenCL (AOCL) compiler to synthesize the benchmark applications, they listed the resource usage and performance of each kernel that can be successfully synthesized by the compiler. In this report, we evaluate the resource usage and performance of the CHO benchmark applications using the Intel FPGA SDK for OpenCL and Nallatech 385A FPGA board that features an Arria 10 FPGA device. The focus of the report is to have a better understanding of the resource usage and performance of the kernel implementations using Arria-10 FPGA devices compared to Stratix-5 FPGA devices. In addition, we also gain knowledge about the limitations of the current compiler when it fails to synthesize a benchmark application.

97 MATHEMATICS AND COMPUTING↗

OpenCL: Free Your GPU... and the rest of your system too!

For the uninitiated, OpenCL may just seem like a tool for accessing GPUs in a portable manner, an alternative to CUDA that works for ATI cards. This view of OpenCL is a vast oversimplification. The truth is that OpenCL offers a rich set of features that will provide the infrastructure needed for application developers to fully utilize an increasingly heterogeneous hardware future. In this article, we will consider how OpenCL can address some of the coming challenges, and what the OpenCL community is doing to insure that there is a vibrant ecosystem of support, as more and more users turn to OpenCL from vendor-locked solutions. By now, we’ve all heard of the power wall, and we know that the only way to climb over it is parallelism. Data-level parallelism (think GPUs) was the first technology to make a mark in the new computing age. Employing very uniform work units, data parallel implementations have achieved noteworthy performance gains for many scientific algorithms, and OpenCL certainly has support for this model. However, this is not the end of the story. As CPU vendors seek to adapt their architectures to the new parallel reality (and GPU vendors evolve theirs), alternate ideas about how to expose and exploit parallelism within a program are surfacing. These designs seek to mitigate performance limitations for algorithms where the majority of the work does not fit neatly into non-divergent units of execution.

97 MATHEMATICS AND COMPUTING↗

A Distributed OpenCL Framework using Redundant Computation and Data Replication

Applications written solely in OpenCL or CUDA cannot execute on a cluster as a whole. Most previous approaches that extend these programming models to clusters are based on a common idea: designating a centralized host node and coordinating the other nodes with the host for computation. However, the centralized host node is a serious performance bottleneck when the number of nodes is large. In this paper, we propose a scalable and distributed OpenCL framework called SnuCL-D for large-scale clusters. SnuCL-D's remote device virtualization provides an OpenCL application with an illusion that all compute devices in a cluster are confined in a single node. To reduce the amount of control-message and data communication between nodes, SnuCL-D replicates the OpenCL host program execution and data in each node. We also propose a new OpenCL host API function and a queueing optimization technique that significantly reduce the overhead incurred by the previous centralized approaches. To show the effectiveness of SnuCL-D, we evaluate SnuCL-D with a microbenchmark and eleven benchmark applications on a large-scale CPU cluster and a medium-scale GPU cluster.

Kim, Junghyun↗

Evaluation of the Single-precision Floatingpoint Vector Add Kernel Using the Intel FPGA SDK for OpenCL

Open Computing Language (OpenCL) is a high-level language that enables software programmers to explore Field Programmable Gate Arrays (FPGAs) for application acceleration. The Intel FPGA software development kit (SDK) for OpenCL allows a user to specify applications at a high level and explore the performance of low-level hardware acceleration. In this report, we present the FPGA performance and power consumption results of the single-precision floating-point vector add OpenCL kernel using the Intel FPGA SDK for OpenCL on the Nallatech 385A FPGA board. The board features an Arria 10 FPGA. We evaluate the FPGA implementations using the compute unit duplication and kernel vectorization optimization techniques. On the Nallatech 385A FPGA board, the maximum compute kernel bandwidth we achieve is 25.8 GB/s, approximately 76% of the peak memory bandwidth. The power consumption of the FPGA device when running the kernels ranges from 29W to 42W.

97 MATHEMATICS AND COMPUTING↗

Optimized Data Transfers Based on the OpenCL Event Management Mechanism

In standard OpenCL programming, hosts are supposed to control their compute devices. Since compute devices are dedicated to kernel computation, only hosts can execute several kinds of data transfers such as internode communication and file access. These data transfers require one host to simultaneously play two or more roles due to the need for collaboration between the host and devices. The codes for such data transfers are likely to be system-specific, resulting in low portability. Here, this article proposes an OpenCL extension that incorporates such data transfers into the OpenCL event management mechanism. Unlike the current OpenCL standard, the main thread running on the host is not blocked to serialize dependent operations. Hence, an application can easily use the opportunities to overlap parallel activities of hosts and compute devices. In addition, the implementation details of data transfers are hidden behind the extension, and application programmers can use the optimized data transfers without any tricky programming techniques. The evaluation results show that the proposed extension can use the optimized data transfer implementation and thereby increase the sustained data transfer performance by about 18% for a real application accessing a big data file.

97 MATHEMATICS AND COMPUTING↗

Bringing OpenCL to Commodity RISC-V CPUs

The importance of open-source hardware has been increasing in recent years with the introduction of the RISC-V Open ISA. This has also accelerated the push for support of the open-source software stack from compiler tools to full-blown operating systems. Parallel computing with today’s Application Programming Interfaces such as OpenCL has proven to be effective at leveraging the parallelism in commodity multi-core processors and programmable parallel accelerators. However, to the best of our knowledge, there is currently no publicly available implementation of OpenCL targeting commodity RISC-V processors that is accessible to the open-source community. Besides opening RISC-V to the existing rich variety of scientific parallel applications, OpenCL also provides access to a unique genre of benchmarks useful in computer architecture research. In this work, we extended an Open-source implementation of OpenCL to target RISC-V CPUs. Our work not only cover commodity multi-core RISC-V processors, but also plethora of low- profile embedded RISC-V CPUs that often do not support atomic instructions or multi-threading.

Tine, Blaise↗

Evaluating and Optimizing OpenCL Base64 Data Unpacking Kernel with FPGA

Development of applications using OpenCL targeting FPGAs is an emerging approach on heterogeneous computing systems. This paper uses the data unpacking algorithm in Base64 encoding as a case study to present programming and optimization techniques, and experimental results of the OpenCL-based implementations on an FPGA. We explain the algorithm and evaluate the performance of the kernel implementations with Intel's FPGA OpenCL SDK. The experimental results show kernel vectorization and duplication are two optimization techniques that can improve the kernel performance. The performance of kernel duplication is also closely related to the local work size. Our experiment shows 16-lane vectorization increases the bandwidth by a factor of 2 to 10 for large input data sizes. Moreover, the performance of kernel duplication using 16 compute units is 40% to 1.5% less than that of kernel vectorization depending on the input size. Tuning the local work size can improve the kernel performance by a factor of 3 to 23. For this kernel, using local memory is not an effective technique to improve the kernel performance because input data is not reused. A combination of vectorization and duplication achieves the highest performance of 12.3 GiB/s. Compared to an Intel Xeon E5 CPU and an Nvidia Tesla K80 GPU, the performance of the kernel on the Arria 10 FPGA is 6.7X faster than the CPU and 3X slower than the GPU. The performance per watt on the FPGA is 20.5X higher than the CPU and 1.19X lower than the GPU.

Base64 Encoding↗

Maestro: Data Orchestation and Tuning for OpenCL Devices

Abstract. As heterogeneous computing platforms become more prevalent, the programmer must account for complex memory hierarchies in addition to the diculties of parallel programming. OpenCL is an open standard for parallel computing that helps alleviate this diculty by providing a portable set of abstractions for device memory hierarchies. However, OpenCL requires that the programmer explicitly controls data transfer and device synchronization, two tedious and error-prone tasks. This paper introduces Maestro, an open source library for data orchestration on OpenCL devices. Maestro provides automatic data transfer, task decomposition across multiple devices, and autotuning of dynamic execution parameters for some types of problems.

99 GENERAL AND MISCELLANEOUS↗

SU-E-T-499: Initial Developments of An OpenCL-Based Cross-Platform Monte Carlo Dose Engine for Carbon Ion Therapy

Purpose Dose calculation is of critical importance for carbon ion therapy. Monte Carlo (MC) simulation is considered to be the most accurate method for calculation of absorbed dose and of all the more fundamental physical quantities related to biological effects. The long computation time, however, limits its routine clinical applications. We have recently started developing a fast MC package, gCMC for carbon therapy on a parallel processing platform, e.g. GPU, aiming at achieving sufficient efficiency to enable MC in clinically important tasks. This abstract reports our progress. Methods gCMC was developed in OpenCL environment. Our initial developments focused on water material. gCMC supported carbon ion transport in the energy range of 1–450 MeV/u. A Class II condensed history algorithm was implemented for charged particle transport simulations with stopping power computed via Bethe-Bloch equation. Energy straggling and multiple scattering were modeled. Total cross section of nuclear interaction was extracted from Geant4. At present, nuclear interaction events were sampled but transports of secondary particles were not included. Results We tested cases with a homogeneous water phantom and a pencil carbon ion beam with energy of 200–400 MeV/u. When only electro-magnetic channel was included, dose/fluence difference between gCMC and Geant4 results averaged within 10% isodose line was <0.5% of the maximum dose/fluence. After enabling nuclear interactions without transporting secondary particles, dose and fluence agreed with the corresponding results computed by Geant4 with <1% difference. Due to the support for multiple platforms of OpenCL, gCMC was executable on NVidia and AMD GPUs, and Intel CPUs. It took ∼50 sec to transport 107 200MeV/u source carbon ions on an NVidia Titan GPU card. Conclusion Preliminary studies have demonstrated the accuracy and efficiency of gCMC. With further developments in near future, gCMC will potentially achieve clinically acceptable fast and accurate MC simulations for carbon ion therapy.

60 APPLIED LIFE SCIENCES↗

SU-C-BRC-06: OpenCL-Based Cross-Platform Monte Carlo Simulation Package for Carbon Ion Therapy

Purpose: Monte Carlo (MC) simulation is considered to be the most accurate method for calculation of absorbed dose and fundamental physical quantities related to biological effects in carbon ion therapy. Its long computation time impedes clinical and research applications. We have developed an MC package, goCMC, on parallel processing platforms, aiming at achieving accurate and efficient simulations for carbon therapy. Methods: goCMC was developed under OpenCL framework. It supported transport simulation in voxelized geometry with kinetic energy up to 450 MeV/u. Class II condensed history algorithm was employed for charged particle transport with stopping power computed via Bethe-Bloch equation. Secondary electrons were not transported with their energy locally deposited. Energy straggling and multiple scattering were modeled. Production of secondary charged particles from nuclear interactions was implemented based on cross section and yield data from Geant4. They were transported via the condensed history scheme. goCMC supported scoring various quantities of interest e.g. physical dose, particle fluence, spectrum, linear energy transfer, and positron emitting nuclei. Results: goCMC has been benchmarked against Geant4 with different phantoms and beam energies. For 100 MeV/u, 250 MeV/u and 400 MeV/u beams impinging to a water phantom, range difference was 0.03 mm, 0.20 mm and 0.53 mm, and mean dose difference was 0.47%, 0.72% and 0.79%, respectively. goCMC can run on various computing devices. Depending on the beam energy and voxel size, it took 20∼100 seconds to simulate 10{sup 7} carbons on an AMD Radeon GPU card. The corresponding CPU time for Geant4 with the same setup was 60∼100 hours. Conclusion: We have developed an OpenCL-based cross-platform carbon MC simulation package, goCMC. Its accuracy, efficiency and portability make goCMC attractive for research and clinical applications in carbon therapy.

60 APPLIED LIFE SCIENCES↗

Evaluating LULESH Kernels on OpenCL FPGA

FPGAs are becoming promising heterogeneous computing components for high-performance computing. In this paper, we evaluate the resource utilizations, performance, and performance per watt of our implementations of the LULESH kernels in OpenCL on an Arria10-based FPGA platform. LULESH is a complex proxy application in the CORAL benchmark suite. We choose two representative kernels “CalcFBHourglassForceForElems” and “EvalEOSForElems” from the application in our study. Compared with the baseline implementations, our optimizations improve the performance by a factor of 1.65X and 2.96X for the two kernels on the FPGA, respectively. Using directives for accelerator programming, we also evaluate the performance of the kernels on an Intel Xeon 16-core CPU and an Nvidia K80 GPU. We find that the FPGA, constrained by the memory bandwidth, can perform 1.05X to 3.4X better than the CPU and GPU for small problem sizes. For the first kernel, the performance per watt on the FPGA is 1.59X and 7.1X higher than that on an Intel Xeon 16-core CPU and an Nvidia K80 GPU, respectively. For the second kernel, the performance per watt on the GPU is 1.82X higher than that on the FPGA. However, the performance per watt on the FPGA is 1.77X higher than that on the CPU.

FPGA↗

Tausch: A halo exchange library for large heterogeneous computing systems using MPI, OpenCL, and CUDA

Exchanging halo data is a common task in modern scientific computing applications and efficient handling of this operation is critical for the performance of the overall simulation. Tausch is a novel header-only library that provides a simple API for efficiently handling these types of data movements. Tausch supports both simple CPU-only systems, but also more complex heterogeneous systems with both CPUs and GPUs. It currently supports both OpenCL and CUDA for communicating with GPGPU devices, and allows for communication between GPGPUs and CPUs. The API allows for drop-in replacement in existing codes and can be used for the communication layer in new codes. This paper provides an overview of the approach taken in Tausch, and a performance analysis that demonstrates expected and achieved performance. Here, we highlight the ease of use and performance with three applications: First Tausch is compared to the halo exchange framework from two Mantevo applications, HPCCG and miniFE, and then it is used to replace a legacy halo exchange library in the flexible multigrid solver framework Cedar.

97 MATHEMATICS AND COMPUTING↗

Performance Portability Evaluation of OpenCL Benchmarks across Intel and NVIDIA Platforms

We evaluate the capabilities of vendor-provided OpenCL implementations for performance portability across multiple computing platforms. The Rodinia benchmark suite is used for this evaluation. We apply the metric defined by Pennycook et al., and we use roofline efficiency from the Roofline performance model as the "performance efficiency" in the metric's definition. We found that the delivered performance portability is similar for several benchmarks, even if the roofline-based performance efficiencies across platforms are very different among the benchmarks. To help distinguish between these instances, we extend the metric by adding the standard deviation of the performance efficiencies for each benchmark. We argue that the standard deviation gives additional insight into performance portability assessment since it adds the performance variability across platforms. Additionally, we discuss the challenges to measure performance portability associated with algorithms and system software. In terms of algorithms, we need to carefully construct the benchmarks and appropriately use the concurrency available on a platform. In terms of system software, we depend on the vendor performance tools to support the desired programming model and runtime to be able to measure the metrics of interest.

GPU↗

Acceleration of Radiance for Lighting Simulation by Using Parallel Computing with OpenCL

We report on the acceleration of annual daylighting simulations for fenestration systems in the Radiance ray-tracing program. The algorithm was optimized to reduce both the redundant data input/output operations and the floating-point operations. To further accelerate the simulation speed, the calculation for matrix multiplications was implemented using parallel computing on a graphics processing unit. We used OpenCL, which is a cross-platform parallel programming language. Numerical experiments show that the combination of the above measures can speed up the annual daylighting simulations 101.7 times or 28.6 times when the sky vector has 146 or 2306 elements, respectively.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

A hybrid programming model for compressible gas dynamics using openCL

The current trend towards multicore/manycore and accelerated architectures presents challenges, both in portability, and also in the choices that developers must make on how to use the resources that these architectures provide. This paper explores some of the possibilities that are enabled by the Open Computing Language (OpenCL), and proposes a programming model that allows developers and scientists to more fully subscribe hybrid compute nodes, while, at the same time, reducing the impact of system failure.

97 MATHEMATICS AND COMPUTING↗

SU-E-T-112: An OpenCL-Based Cross-Platform Monte Carlo Dose Engine (oclMC) for Coupled Photon-Electron Transport

Purpose: Low computational efficiency of Monte Carlo (MC) dose calculation impedes its clinical applications. Although a number of MC dose packages have been developed over the past few years, enabling fast MC dose calculations, most of these packages were developed under NVidia’s CUDA environment. This limited their code portability to other platforms, hindering the introduction of GPU-based MC dose engines to clinical practice. To solve this problem, we developed a cross-platform fast MC dose engine named oclMC under OpenCL environment for external photon and electron radiotherapy. Methods: Coupled photon-electron simulation was implemented with standard analogue simulation scheme for photon transport and Class II condensed history scheme for electron transport. We tested the accuracy and efficiency of oclMC by comparing the doses calculated using oclMC and gDPM, a previously developed GPU-based MC code on NVidia GPU platform, for a 15MeV electron beam and a 6MV photon beam in a homogenous water phantom, a water-bone-lung-water slab phantom and a half-slab phantom. We also tested code portability of oclMC on different devices, including an NVidia GPU, two AMD GPUs and an Intel CPU. Results: Satisfactory agreements were observed in all photon and electron cases, with ∼0.48%–0.53% average dose differences at regions within 10% isodose line for electron beam cases and ∼0.15%–0.17% for photon beam cases. It took oclMC 3–4 sec to perform transport simulation for electron beam on NVidia Titan GPU and 35–51 sec for photon beam, both with ∼0.5% statistical uncertainty. The computation was 6%–17% slower than gDPM due to the differences in both physics model and development environment, which is considered not significant for clinical applications. In terms of code portability, gDPM only runs on NVidia GPUs, while oclMC successfully runs on all the tested devices. Conclusion: oclMC is an accurate and fast MC dose engine. Its high cross-platform portability makes it clinically attractive.

60 APPLIED LIFE SCIENCES↗

TU-AB-BRC-02: Accuracy Evaluation of GPU-Based OpenCL Carbon Monte Carlo Package (goCMC) in Biological Dose and Microdosimetry in Comparison to FLUKA Simulations

Purpose: One of the most accurate methods for radiation transport is Monte Carlo (MC) simulation. Long computation time prevents its wide applications in clinic. We have recently developed a fast MC code for carbon ion therapy called GPU-based OpenCL Carbon Monte Carlo (goCMC) and its accuracy in physical dose has been established. Since radiobiology is an indispensible aspect of carbon ion therapy, this study evaluates accuracy of goCMC in biological dose and microdosimetry by benchmarking it with FLUKA. Methods: We performed simulations of a carbon pencil beam with 150, 300 and 450 MeV/u in a homogeneous water phantom using goCMC and FLUKA. Dose and energy spectra for primary and secondary ions on the central beam axis were recorded. Repair-misrepair-fixation model was employed to calculate Relative Biological Effectiveness (RBE). Monte Carlo Damage Simulation (MCDS) tool was used to calculate microdosimetry parameters. Results: Physical dose differences on the central axis were <1.6% of the maximum value. Before the Bragg peak, differences in RBE and RBE-weighted dose were <2% and <1%. At the Bragg peak, the differences were 12.5% caused by small range discrepancy and sensitivity of RBE to beam spectra. Consequently, RBE-weighted dose difference was 11%. Beyond the peak, RBE differences were <20% and primarily caused by differences in the Helium-4 spectrum. However, the RBE-weighted dose agreed within 1% due to the low physical dose. Differences in microdosimetric quantities were small except at the Bragg peak. The simulation time per source particle with FLUKA was 0.08 sec, while goCMC was approximately 1000 times faster. Conclusion: Physical doses computed by FLUKA and goCMC were in good agreement. Although relatively large RBE differences were observed at and beyond the Bragg peak, the RBE-weighted dose differences were considered to be acceptable.

60 APPLIED LIFE SCIENCES↗